I've been experimenting more and want to point out how fantastic the 'Film Desat' is. I find pleasing desaturation a lot harder to achieve than pleasing saturation and yours is one of the rare tools that gets it right for my taste.
2499 DRT - an alternative picture formation pipeline
- Thread starter Juan Pablo Zambrano
- Start date
I am glad , film desat is one of the thing I am most proud of , there is so much work on saturation options but not desaturation.
Juan, sent you a DM, please check.
Hello, Juan!
Are you planning to add JPLog2 as a color space? This is taking off, and now I am grading exclusively using JOLog. Offering such options will expand the use of your DCTL. I can provide more information on it, if you don’t have it (which I doubt since it is public and free).
Thanks,
Luid
Are you planning to add JPLog2 as a color space? This is taking off, and now I am grading exclusively using JOLog. Offering such options will expand the use of your DCTL. I can provide more information on it, if you don’t have it (which I doubt since it is public and free).
Thanks,
Luid
Does it take off? I’m still unsure how much of JPLog2 reflects genuine industry needs versus brand-driven promotion. From my perspective it seems more like a proprietary approach that benefits Colourlab.ai’s ecosystem and brand awareness rather than offering advantages over existing options - unless maybe when being tied to Colourlab's own tool chain and formats.
I don't mean to start a discussion just wanted to add that I personally think 2499 does not need to address it.
I don't mean to start a discussion just wanted to add that I personally think 2499 does not need to address it.
Hi Luis,
I'm very unlikely to add JPLOG2 support to 2499. I have several concerns about this Intermediate Space:
First, I disagree with JPLOG2's fundamental design approach. It compresses too much dynamic range into a limited space, which causes problems with standard tools that aren't color space aware. This compression also makes traditional controls like lift, gamma, gain, and offset less effective since their impact on the image becomes too similar.
Second, JPLOG2 uses ACES AP1 primaries, which I believe are too restrictive for most cameras. I think having adequate primary coverage is more critical than maximizing dynamic range encoding.
It's worth noting that while JPLOG2 was initially considered for ACES 2.0, it was ultimately deemed unnecessary, a decision I support.
Overall, I see more drawbacks than advantages with JPLOG2, and I haven't encountered compelling use cases that would justify it as a working color space.
something like the following could work if you really wanted to grade in jplog under the 2499 drt...
camera color space > JPLOG2 > ACES2065 > 2499 r709
or
camera color space > JPLOG2 > ACES2065 > 2499 log > 2499 r709
I've had good luck/successes with these approaches, however I've only been happy to use JPLog2 on Alexa 35 footage, owing to its borderline-unusable (esp. down in the toe), very low mid-grey (in my opinion).
Also tried (in Resolve) DWG/DI and ACEScct but both feel twitchy/jumpy with Alexa 35 sources (below mid-grey). JPLog2 makes linear gain, and linear offset, feel pretty nice here so that's what I've used it on. Seems like for other sources DWG/DI, or ACEScct, are 'fine' and JPLog2 isn't needed. Admire the design approach though: sane/sensible middle gray point and evenly apportioned code values per stop. Main downside, for me, is the low-ish diffuse white means every shot needs the highlights touched up (even if otherwise 'perfect'). Debatable whether this is worth it.
Back on topic, still would love to see ACEScct added as a valid input to JP_2499, both for 'completeness' sake and, selfishly, because I often do lookdev in cct (have many PowerGrades, etc. that have been made for it) and would rather not do yet another conversion prior to (and only for) the DRT. I've read through the thread so know this isn't planned, just underlining that a some-sized part of the community would like to see this.
In any case, thanks for making such a lovely DRT! We live in a new age now, with tools like this; love to see it.
Antigravity released a new drone and when reading through the press release, for a confused second I thought it records 2499 LOG 
Hey Juan! Thank you so much for this useful DRT.
When you say film desat — is that just a feature of 2499? Or is that something I should be adjusting?
Thanks!
Henri
Hello Henri, Film Desat is one of the controls in the Custom Curves for 2499 log tool, it's one of the dctles available along with 2499. https://github.com/JuanPabloZambrano/DCTL/blob/main/2499_DRT/CustomCurves for LOG2499.dctle
Am I overlooking something or is the source code not available anymore? The DCTLs are all encrypted now?
Split tone presets for 2499LOG - black dots artifacts. The only DСTL that produces artifacts.
Attachments
Last edited:
There are plenty of DCTLs that create artifacts. Try building a node tree with more serial nodes and less parallel nodes. Very often, simply reorganizing your node tree is enough to make the artifacts disappear. Any way, I’ve already achieved some beautiful color gradings for my shorts using the Split Tone presets and two 2499 DRTs and others DCTLs (without artifact)
I don't use parallel nodes at all.
This simple node tree works perfectly for me (Mac OS 26.2; Resolve 20.3.2)
Already same pipe but in groups
Attachments
Why are you applying tone mapping in the first CST? Normally, when converting from a wide color space like SLOG3 to DWG, there's no need for tone mapping.
Does this make a difference?
Attachments
So I don't understand.
